Data is everywhere and we are creating more and more every second. The graphic below illustrates the data explosion. These numbers clearly show that the market and your customers are moving faster, generating increasing volumes of data , and expecting more from you in return.
To meet customer expectations, companies of all sizes are investing heavily in capturing and analyzing data. Investing alone is not enough: if the data isn’t clean and accurate, it’s money down the drain. Poor quality data creates a domino effect of disaster across an organization, affecting C-Level, Sales, Marketing, Support, IT, Finance, Legal and every other function you can think of. Data quality issues can lead to both hard and soft costs for your company including -- but not limited to -- the following:
- Lost revenues
- Business inefficiencies
- Compliance or regulatory fines
- Brand or credibility damage
- Failed marketing initiatives
At Summa, we worked with a client in the home delivery pharmacy space. The team determined that this client lost an average of $1000 per bad address in their system. This bad data had administrative costs due to the returned medication, product costs due to the forced discarding of the returned medication, and customer health costs due to the additional patient risk associated with medications not arriving on time.
Preventing or reining-in these costs requires data governance, in other words, identifying and upholding data definition, data integrity and related standards.
Here are five essential steps in getting started on a data governance initiative that will achieve the results you want.
1. Understand your legal and regulatory environment.
What governmental or regulatory bodies and policies must you adhere to? These include the legal requirements around data encryption, data retention, data auditing, and data residency. (Health Insurance Portability and Accountability Act [HIPAA], Sarbanes-Oxley [SOX], European Union [EU], etc)
Typically these policies and requirements are well defined by entities outside of your business. There are well established pathways to work within or around. To understand the full picture, you will probably need to involve many additional teams/departments in your company such as Disaster Recovery, Security, Data Backup, Auditing, and Change Management.
2. Identify your requirements for data accuracy, precision, grain, and quality.
Review your data objects and all of the fields/properties. Once you know which data elements are most important to the teams that use the data, you can identify the gaps between the current state and the improved future state.
Decisions made here can have lasting effects. Be sure to involve the correct resources and subject matter experts when determining these guidelines. The team must define up-front the optimal level of precision and grain, and bring insight into future needs. It is very costly to modify your data to a finer grained or more precise level after the initial implementation. In comparison, it is relatively inexpensive to pivot to a coarser grained or less precise level of detail after the system is live.
3. Plot your data flow, where it is gathered, and how it is used to further your organizational strategy.
The most effective exercise is to review, refine and update existing systems architecture and integration diagrams and documentation. What systems are storing and providing your data and what are the endpoints and connections between systems? Once you can see the data flow, you can identify the data entry points for which you will need to implement and enforce governance policies and procedures.
4. Standardize your terminology and data classification system.
Large organizations often have siloed departments or teams that use the same data in different ways and have developed their own terminology and classifications for it. This lack of consistency across the organization can be a costly challenge when defining and implementing data governance. You will need to bring these groups together to give them the new common terminology or (better) enlist them in the process to develop the new data dictionary, so to speak. Once established, this terminology and classification system can simplify the data governance policies and improve adoption.
5. Assemble a balanced team.
We are assuming you have already received executive level buy-in for the initiative and have acquired funding. Given the organization-wide impact of this initiative, your stakeholders will represent all systems for which data is captured, consumed, stored and/or exposed. Team members should include:
- Representatives of groups who create and use data
- Technical resources like data architects, DBAs, and developer teams
- Champions who will help with long term adoption
- Detractors who will have their concerns publicized early and will contribute to the development of policies which can have a positive effect on all
Launching your data governance program is just the beginning. Once it is under way, your to-do list will include establishing data stewardship to monitor day to day activities, identifying the best tools for recognizing and resolving data issues, and measuring the success of the program.
Read more about our healthcare solutions:
- Maximizing the Capabilities of Health Cloud
- I'm the Patient, Remember Me?
- Interoperability & Quality Process Supporting EHRs
- Is Salesforce HIPAA Compliant?
